Dev

dev@lists.roundcube.net

4296 discussions

Purpose of change_identity JavaScript event trigger? (was: [Svn] [roundcube/roundcubemail] af61b9: Adds change_identity JavaScript event trigger)
by Michael Heydekamp 04 Nov '13

04 Nov '13

Hi there, may I ask what the intention of the commit quoted below exactly is...? Here it changed the behaviour of select_identity apparently in so far as the address in the From: header is now being selected as the identity, even if it does not match the full mailbox spec (name+address). This is in contradiction to what Alec recently wrote: ---------------------------------------------------------------------------- Am 14.10.2013 11:03, schrieb A.L.E.C: > On 10/14/2013 10:47 AM, Michael Heydekamp wrote: > >> Thanks. But are you sure about "From"? During my tests I didn't manage to >> select the identity in the From: header at all. The default identity had >> been selected instead (unless From: and Return-path: were identical). > > Well, From header check matches full mailbox spec., e.g. "Roundcube > Users <dev(a)lists.roundcube.net>". > > This is because we want to handle correctly situation when user has many > identities with same address but different name. It's not perfect and > I'm not saying we don't need an improvement here. ---------------------------------------------------------------------------- And an hour later, he wrote: ---------------------------------------------------------------------------- Am 14.10.2013 12:03, schrieb A.L.E.C: > For To/Cc the matching is different. We check addresses, but give bigger > prio to identity where address+name matches. So, here we're safe. > I suppose we could do the same for From with some more coding. Feel free > to create a ticket. ---------------------------------------------------------------------------- So I created a ticket: http://trac.roundcube.net/ticket/1489378 Is the commit below now the reaction to this ticket? Most likely not, as the ticket hasn't been closed yet. But the commit serves at least a part of the purpose of this ticket (= use From: header as identity even if it doesn't match full mailbox spec). So I'm not sure what's this all about, and if the result that I'm seeing is indeed intentional. Or am I wrong and the behaviour described above is related to a totally different commit...? Cheers, Michael (confused) -- Michael Heydekamp Co-Admin freexp.de Düsseldorf/Germany Am 18.10.2013 21:40, schrieb GitHub: > Branch: refs/heads/master > Home: https://github.com/roundcube/roundcubemail > Commit: af61b9705e40efa938bd38d95173a41cf210e6dc > > https://github.com/roundcube/roundcubemail/commit/af61b9705e40efa938bd38d95… > Author: Niklas Femerstrand <nik(a)qnrq.se> > Date: 2013-10-18 (Fri, 18 Oct 2013) > > Changed paths: > M program/js/app.js > > Log Message: > ----------- > Adds change_identity JavaScript event trigger > > > Commit: 771a82df9031485b3f9e7decf6b263c6847507b1 > > https://github.com/roundcube/roundcubemail/commit/771a82df9031485b3f9e7decf… > Author: Thomas B. <thomas(a)roundcube.net> > Date: 2013-10-18 (Fri, 18 Oct 2013) > > Changed paths: > M program/js/app.js > > Log Message: > ----------- > Merge pull request #139 from qnrq/master > > Adds change_identity JavaScript event trigger > > > Compare: > https://github.com/roundcube/roundcubemail/compare/1498ac81c946...771a82df9… > > _______________________________________________ > Roundcube SVN commits mailing list > svn(a)lists.roundcube.net > http://lists.roundcube.net/mailman/listinfo/svn

3 3

Plugins, $config and RC 1.0
by Michael Heydekamp 04 Nov '13

04 Nov '13

Hi there, we're unsure and a bit confused about how to handle the plugin configs in our 1.0-git test environment: 1) Almost all plugins do still use the legacy "$rcmail_config" array in their .dist file, rather than "$config". Do we need to manually run a search/replace across all config.inc.php(.dist)? 2) The plugin 'password' for instance does already use the "$config" array in it's config.php.dist. But this is not being reflected in config.inc.php (shouldn't there be some sort of automatism?). 3) Furthermore, the same plugin 'password' does contain a number of new config settings in .dist - but also those did not find their way into the .php file. So we have to merge them into .php as well? If so: Hmmm, I have seen (and even developed) more convenient and automated ways of updating configs already, but that was decades ago under DOS/16... ;) JFTR: The 'password' plugin is just an arbitrary example, the same may/will most likely apply to other plugins as well. But 'password' is just the first one where we recognized that "$config" is being used at least in the .dist file already. Any helpful advice appreciated. Cheers, -- Michael Heydekamp Co-Admin freexp.de Düsseldorf/Germany

3 12

Regression upon right-clicking on message
by Michael Heydekamp 04 Nov '13

04 Nov '13

See attached: In 0.9.5, a sub-menu opens upon right-clicking and hovering the mouse on the "Mehr..." item. In 1.0-git, no such sub-menu opens. But it did still open a few days ago. <sigh> (Win7/64, IE10) Cheers, -- Michael Heydekamp Co-Admin freexp.de Düsseldorf/Germany

3 3

state of unit tests
by Charles McNulty 31 Oct '13

31 Oct '13

What is the current state of the unit tests? I am getting many failures of both the PHP and Selenium tests, but I was running them off on master, not a released version. Part of the problem was that I was hitting the 120 second PHP timeout that's hardcoded into iniset.php. When you run the tests do you manually change that or is there some other way of running the tests that bypasses the execution timeout? Are tests being added to and maintained? I've become used to test-driven development with jQuery, and it feels a little dirty to submit a bug or a pull request without a corresponding unit test. -Charles -- Sent using Roundcube http://roundcube.net/

2 3

focused messages - dotted outline
by Charles McNulty 30 Oct '13

30 Oct '13

Hello team, I spent quite a few hours back in 2006 "scratching my own itch" in the roundcube code. I haven't had much time to work on it since then, but I happen to have some time at the moment to dedicate to the project. I just submitted a pull request that will make "control-click" and "control-arrow-key" behave more like a native application. One thing that has always bothered my about the roundcube layout is that there's no way to know which message in the list has the "focus" in cases where the focus is separate from the selected message. For instance if you select a message, and then CTRL-click that same message, it will be unselected, but it still has the focus, because if you then click the down arrow it will select the next message. The css "outline" attribute would work for this, but there are some inconsistencies with where exactly each browser places the outline - and IE (including IE10) does not support the outline-offset attribute at all, which means that the outline doesn't work very well in IE (no surprise). To see what I mean, you can insert the following style into styles.css: .records-table tr.focused { outline:1px dotted black; outline-offset: -1px; } I'd appreciate any thoughts on the issue! -Charles McNulty

3 7

keyboard documentation and further ideas
by Charles McNulty 30 Oct '13

30 Oct '13

I scoured the code for keyboard handling and here's what I found: Message & Contact List (unless otherwise noted): Up and down arrow keys: Select next/previous item When selecting a message in a thread: Plus/minus keys: Open/close the thread Spacebar: Toggles thread open and closed +/= key: Open thread (firefox - not chrome) Page-up/page-down: Load previous/next page Home/end: Select first/last item Enter key: Show message (not applicable to contacts) Escape key: Cancel mouse drag (mid-drag) Delete key: Delete selected items Backspace key: Delete selected messages only (does not work for contacts) It doesn't look like this is documented anywhere. Anyone care to add this to the wiki? I'd be happy to create/manage a wiki page for key-bindings, or it could just be added to the User_Guide page. Personally, I think they could be tweaked though, and maybe that should happen before documenting the behavior. 1) The (+/=) key support is browser dependent, and could probably just be dropped. 2) The backspace key works for messages but not contacts and seems redundant with the delete key - additionally some people use the backspace key for "back" in a browser. I'm not suggesting we support using the backspace key for browser-back, but I am suggesting that we might not want to delete messages if a user presses it intending to go back. 3) I think that we should add left/right cursor (arrow) key support to expand/collapse threads. -Charlie

6 6

problem latest git and mobile devices
by Cor Bosman 29 Oct '13

29 Oct '13

Hey all, ive gotten some complaints regarding a recent git version with a test user base. It seems you can no longer scroll a message in the messageview. I can reproduce this on my iphone with today's git version but ive also had android complaints. This is a pretty serious issue as it makes rc unusable on mobile devices. Reports show that that is on average about 20% of a site's userbase. Cor

5 12

Re: [RCD] [RCU] roundcube 0.9.3 - Invalid request. no data saved
by Reindl Harald 29 Oct '13

29 Oct '13

Am 29.10.2013 14:53, schrieb Igor Zinovik: > My roundcube does not allow me to login into my mailbox. It just writes: > " invalid request no data was saved " > > I enabled all logging settings in main.inc.php > $rcmail_config['smtp_log'] = true; > $rcmail_config['log_logins'] = true; > $rcmail_config['log_session'] = true; > $rcmail_config['sql_debug'] = true; > $rcmail_config['imap_debug'] = true; > > My Apache is able to successfully save sessions to /var/lib/php5/ that's nice but roundcube still insists to not use the php session-handler otherwise i would no longer need to carry the patch from months ago in the RPM-SPEC Patch2: roundcubemail_native_session_handler.patch [builduser@buildserver:/rpmbuild/SPECS]$ cat ../SOURCES/roundcubemail_native_session_handler.patch diff --git a/config/main.inc.php.dist b/config/main.inc.php.dist index 4a73ff1..58332cb 100644 --- a/config/main.inc.php.dist +++ b/config/main.inc.php.dist @@ -254,9 +254,10 @@ $rcmail_config['session_name'] = null; // Session path. Defaults to PHP session.cookie_path setting. $rcmail_config['session_path'] = null; -// Backend to use for session storage. Can either be 'db' (default) or 'memcache' -// If set to memcache, a list of servers need to be specified in 'memcache_hosts' +// Backend to use for session storage. Can either be 'db' (default), 'memcache' or 'php' +// If set to 'memcache', a list of servers need to be specified in 'memcache_hosts' // Make sure the Memcache extension (http://pecl.php.net/package/memcache) version >= 2.0.0 is installed +// Setting this value to 'php' will use the default session save handler configured in PHP $rcmail_config['session_storage'] = 'db'; // Use these hosts for accessing memcached diff --git a/program/include/rcmail.php b/program/include/rcmail.php index 89f2f96..a7ddb1e 100644 --- a/program/include/rcmail.php +++ b/program/include/rcmail.php @@ -746,7 +746,7 @@ class rcmail extends rcube // before closing the database connection, write session data if ($_SERVER['REMOTE_ADDR'] && is_object($this->session)) { - session_write_close(); + $this->session->write_close(); } // write performance stats to logs/console diff --git a/program/lib/Roundcube/rcube.php b/program/lib/Roundcube/rcube.php index 4471ace..4e3f8fc 100644 --- a/program/lib/Roundcube/rcube.php +++ b/program/lib/Roundcube/rcube.php @@ -457,7 +457,6 @@ class rcube ini_set('session.name', $sess_name ? $sess_name : 'roundcube_sessid'); ini_set('session.use_cookies', 1); ini_set('session.use_only_cookies', 1); - ini_set('session.serialize_handler', 'php'); ini_set('session.cookie_httponly', 1); // use database for storing session data @@ -471,7 +470,7 @@ class rcube // start PHP session (if not in CLI mode) if ($_SERVER['REMOTE_ADDR']) { - session_start(); + $this->session->start(); } } diff --git a/program/lib/Roundcube/rcube_session.php b/program/lib/Roundcube/rcube_session.php index dedde22..4e06827 100644 --- a/program/lib/Roundcube/rcube_session.php +++ b/program/lib/Roundcube/rcube_session.php @@ -42,6 +42,7 @@ class rcube_session private $secret = ''; private $ip_check = false; private $logging = false; + private $storage; private $memcache; @@ -59,11 +60,14 @@ class rcube_session $this->set_lifetime($lifetime); // use memcache backend - if ($config->get('session_storage', 'db') == 'memcache') { + $this->storage = $config->get('session_storage', 'db'); + if ($this->storage == 'memcache') { $this->memcache = rcube::get_instance()->get_memcache(); // set custom functions for PHP session management if memcache is available if ($this->memcache) { + ini_set('session.serialize_handler', 'php'); + session_set_save_handler( array($this, 'open'), array($this, 'close'), @@ -79,7 +83,9 @@ class rcube_session true, true); } } - else { + else if ($this->storage != 'php') { + ini_set('session.serialize_handler', 'php'); + // set custom functions for PHP session management session_set_save_handler( array($this, 'open'), @@ -92,6 +98,22 @@ class rcube_session } + /** + * Wrapper for session_start() + */ + public function start() + { + session_start(); + + // copy some session properties to object vars + if ($this->storage == 'php') { + $this->key = session_id(); + $this->ip = $_SESSION['__IP']; + $this->changed = $_SESSION['__MTIME']; + } + } + + public function open($save_path, $session_name) { return true; @@ -116,6 +138,20 @@ class rcube_session /** + * Wrapper for session_write_close() + */ + public function write_close() + { + if ($this->storage == 'php') { + $_SESSION['__IP'] = $this->ip; + $_SESSION['__MTIME'] = time(); + } + + session_write_close(); + } + + + /** * Read session data from database * * @param string Session ID

1 0

Post hooks for DB operations
by Gus 29 Oct '13

29 Oct '13

Hi all, I'm currently writing a plugin that needs to act after an identity has been inserted into database. There is an identity_create hook that is called before the identity is inserted. How about adding a hook (whose argument would be data for the newly created identity, including the identity_id), that would be called after the database record has been created? That way you could write a plugin that could, e.g. 1. use identity_create to start a DB transaction and possibly modify identity data 2. use the new identity_stored (or something similar) hook to write it's own record into its own database table (with a constraint referencing the new identity) and commit the transaction (or rollback if something went wrong) Similar to those, I would also propose post-insert/update hooks for other events that could prove useful, so: identity_stored ( $record ) for after an identity was stored (created or updated) user_created ( $record ) for after a new user has been created contact_stored ( $record ) for after a contact has been stored (created or updated) group_stored ( $record ) for after a group has been stored (created, updated, renamed, members added, members deleted) preferences_stored ( $record ) for after user's preferences have been saved Is this acceptable? It's not too complicated to add those and I'd be willing to contribute the code and document the changes on the wiki. If not, I'd appreciate if you could clue me into how I might be able to write a plugin that needs to do something similar as described above (i.e. create records in additional tables that keep referential integrity with records in Roundcube's own tables). Regards, Tin

1 0

feedback regarding larry
by Cor Bosman 24 Oct '13

24 Oct '13

Hey all, ive been getting some feedback regarding the larry theme. Most of it i can handle myself, but one specifically keeps coming back. People dont realise that this button does anything, and miss all the feature that are accessed through it: http://grab.by/rkkc It's in a bit of a strange place. You click on the left, but it opens something underneath. It doesnt really look very clear what it's supposed to do. I would actually prefer to start it out in open, and let people figure out how to close it, that would sure save us a lot of support calls. Would be even better if it could be more intuitive about that it's supposed to do. Cor

5 8

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

Dev